JP5347909B2 - Example device, program, and method of word or combination of words in patent document - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To present a concrete example assumed by a word or the combination of words used in a claim even if the concrete example assumed by a word or the combination of words used in the claim does not exist in a patent document, concerning a technology for supporting the understanding of a word or the combination of words in the patent document. <P>SOLUTION: An illustrating device includes: a patent document retrieving part 102 for retrieving text data of a description designated by a patent database 101; a rejection reason announcement retrieving part 104 for retrieving text data of a document indicating the reason of a rejection indicated with respect to the description; and a language pair extracting part 106 for extracting from the text data of the document indicating the reason of the rejection, a word or the combination character string of words described in the claim of the description to be an object indicating the concrete example of the word or combination of words and also a word or the combination character string of words indicated in contrast with the word or the combination of words. Extracted data of the pairs of the words or the combinations of words are displayed in a display part 109 by contrast. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to a technology for supporting understanding of words in patent documents, for example, claims.

Patent documents, such as claims, must be read by non-patent specialists for the purpose of inventors to avoid infringement of other companies' patents.
However, the claims described in the claims of patent documents are generally written in words or combinations of words, so that words having a specific meaning are often written as abstract words. It is difficult for an inventor who is unfamiliar with patent documents to specifically understand the invention described in the claims.

Therefore, there is a demand for assistive technology that makes it easy to understand the claims by displaying specific examples assumed by the words used in the claims, thereby facilitating the understanding of the claims.
For example, in the example of claim 1 in the patent document shown in FIG. 1A, the words “content information”, “disclosure unit of individual opinion”, “object of opinion”, “individual evaluation” as indicated by the underlined portion Is used. In this case, by presenting words indicating specific examples of these words, the contents represented by the claims can be imaged and readability can be improved.

As a first conventional technique for supporting the understanding of words in claims, a technique for displaying more specific words and related words from a thesaurus or a synonym dictionary is known.
In addition, as a second conventional technique for supporting the understanding of the words of the claims, the claims are supported by associating the claims with the embodiments or embodiments for carrying out the invention by linking or the like. Technology is known. (For example, Patent Document 1)
In addition, there is also a method for determining a reference reason for a reference document from a result of text structure analysis, and outputting reference relationship information including the reference reason. (For example, Patent Document 2)

JP 2002-149704 A JP 2000-357170 A

  However, in the first prior art described above, the words written in the claims are abstract and may be interpreted differently depending on the field. For this reason, a general thesaurus subordinate has a problem that it is not always an example intended by the inventor.

  Further, the second prior art described above has a problem that it is not possible to effectively support the interpretation of a claim unless an explicit example is written in one patent. It was. For example, in the example shown in FIG. 1B, the term “individual evaluation” is used in claim 1. However, looking at [Disclosure of the Invention] which is a part of the detailed description of the invention, there is only a description of “individual evaluation (for example, good evaluation or bad evaluation)”. For this reason, it is difficult to grasp a specific image in such an example.

  Further, even if specific examples of words in the claims are written in the detailed description of the invention, it is possible to identify the embodiments corresponding to the respective claims or the modes for carrying out the invention. There are various ways of writing specific examples. For this reason, the second prior art described above has a problem that it is difficult to perform a fine association such as a relationship between words mechanically.

  Therefore, in the subject of the present invention, even if there is no specific example that the word used in the claim assumes in the patent document, a specific example that the word used in the claim assumes is presented. It is possible.

  According to an example device for words (words or combinations of words) in a patent document according to one aspect of the present invention, a specific example of words included in a patent document in an apparatus for supporting understanding of words in a patent document A specification that is a target of a specific example of a word from a reception unit that receives information specifying a specification that is a target to indicate, and a rejection reason database that records a document that indicates the reason for rejection indicated for the specification The document search section showing the reason for refusal that searches the text data of the document showing the reason for refusal based on the information specifying the specification, and the text of the document showing the reason for refusal searched. A character string having the same pair expression pattern as the previously stored pair expression pattern is detected from the data, and based on the character string of the pair expression pattern, the character string described in the patent document and the character string are matched. By detecting a pair of character strings each including the character string shown in the above, and deleting unnecessary character strings from the paired character strings, An extraction unit that extracts a character string of a written word, a pair of character strings of a specific example shown in contrast to the word, and an output unit that outputs a set of character strings of the extracted paired words It has.

  Even if the specification of the subject patent does not clearly show an example, the words in the patent document, for example, the claims are made concrete by using the interpretation of the examiner who seems to have a deep understanding of the content of the patent. It is possible to know what is represented in As a result, even a user who is unfamiliar with the description method of a patent document such as a claim can assist in understanding the patent document such as a claim.

It is a figure which shows the example of a description of a claim. It is explanatory drawing of the problem of a prior art. It is a block diagram of 1st Embodiment. It is explanatory drawing of 1st Embodiment. It is a figure which shows the pattern of notification of reasons for refusal. It is a flowchart of a 1st embodiment. It is explanatory drawing of step S403 of FIG. 5 is a detailed flowchart of step S404 (processing for extracting word pairs) in FIG. It is a figure which shows the example of an expression pattern dictionary. It is operation | movement explanatory drawing of 1st Embodiment corresponding to Example 1 of Fig.3 (a). It is operation | movement explanatory drawing of 1st Embodiment corresponding to the example 2 of FIG.3 (b). It is operation | movement explanatory drawing of 1st Embodiment corresponding to Example 3 of FIG.3 (c). It is operation | movement explanatory drawing of 1st Embodiment corresponding to Example 4 (exception rule) of FIG.3 (d). It is a figure which shows the example of a display of a display part. It is a figure which shows the example of the patent database 101. FIG. It is a figure which shows the example of the reason database 105 for refusal. It is a detailed flowchart which shows the specific operation | movement of a pattern matching process. It is a flowchart which shows the unnecessary character deletion process by dependency analysis. It is explanatory drawing of a dependency analysis process. It is a block diagram of 2nd Embodiment. It is a flowchart of a 2nd embodiment. FIG. 16 is a detailed flowchart (part 1) of step S1501 in FIG. 15 (processing for determining which is a specific example); FIG. 17 is a detailed flowchart of step S1601 (process of method 1) in FIG. 6 is a diagram illustrating an example of contents of a word pair table in Method 1. FIG. FIG. 17 is a detailed flowchart of step S1603 (process of method 2) in FIG. 10 is a diagram illustrating an example of contents of a word pair table in Method 2. FIG. FIG. 16 is a detailed flowchart (part 2) of step S1501 in FIG. 15 (processing for determining which is a specific example); FIG. 22 is a detailed flowchart of the processing in step S2103 in FIG. 21 (processing for analyzing the correspondence between claims and examples). It is a figure which shows the example of a word pair database. It is a block diagram (the 1) of 3rd Embodiment. It is a block diagram (the 2) of 3rd Embodiment. It is a flowchart (the 1) of a 3rd embodiment. It is a flowchart (the 2) of a 3rd embodiment. It is a hardware block diagram which implement | achieves the system of embodiment.

Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings.
FIG. 2A is a block diagram of a first embodiment of an example device for words (words or combinations of words) in a patent document. The system according to the first embodiment includes a patent database 101, a patent document search unit 102, an application number extraction unit 103, a rejection reason notification search unit 104, a rejection reason database 105, a word pair extraction unit 106, an expression pattern dictionary 107, and a specific example. An addition target specifying unit 108 and a display unit 109 are provided.

FIG. 2B is an explanatory diagram of a basic operation of the first embodiment.
In the first embodiment, a correspondence relationship between words written in a document indicating a reason for refusal, such as a notice of reasons for refusal, is used for a patent document searched from the patent database 101 (FIG. 1). Specific examples written in other patents are extracted. For example, the reason for refusal 202 for the publication data 201 shown in FIG. 2B corresponding to the patent document will be described. The reason for rejection 202 corresponding to the publication data 201 can be obtained by searching the rejection reason database 105 (FIG. 2A) by the application number described in the “application number” item in the publication data 201. “Individual evaluation” is described in the reason for refusal 202 by using the expression “The slander and the rumor described in the cited reference 2 corresponds to“ individual evaluation ”of the invention of claim 1”. As for a specific example of the above, it is possible to obtain the information that it is “defamatory or rude”. As a result, a display as shown by 203 in FIG. 2B is performed.

Here, “word” refers to a word or a combination of words, and this word or combination of words includes a compound word, a dependency word, a combination of multiple phrases, and the like.
Here, the document indicating the reason for refusal, such as a notice of reasons for refusal, is written by the examiner, so it has a better shape than the gazette and is relatively easy to handle. I can't extract word pairs. There are four writing patterns (writing methods) as shown in Examples 1, 2, 3, and 4 in FIGS.

Example 1 in FIG. 3A is a pattern in which the word “communication state information” of the present application is written before “equivalent” indicating that the word “network topology” in the cited document corresponds to the word “network topology” in parentheses.
Example 2 in FIG. 3B is a pattern in which the words corresponding to the quoted word “bulletin board” and the word “content” of the present application are written in the order of “equivalent” indicating corresponding.

  In Example 3 of FIG. 3C, the words “client computer, A bookstore, convenience chain” of the cited document are collectively written. Then, the corresponding words “first information processing device, third information processing device, second information processing device” of the present application are written in the corresponding order, and “corresponding” indicating the correspondence is indicated after the words of the present application. It is a writing pattern.

Example 4 in FIG. 3D is a pattern in which the word “meta-level task” of the present application is written first and “corresponding” indicating that it corresponds, followed by the word “movement instruction” of the cited document. .
In Examples 1 to 4 above, the words of the present application, the words of cited references, and the like are stored as character string data (hereinafter simply referred to as “character strings”) in the work memory, and are written as the storage positions of the character strings. A pattern or descriptive pattern is determined. In all of Examples 1 to 3, the words of the cited document appear before the words of the present application. In Example 4, however, there are exceptional patterns of writing in which the words of the present application appear before the words of the cited document. That is, when text data of the pattern to be written is developed in the work memory and sequentially scanned, in Examples 1 to 3, scanning is performed in the order of the present application following the cited document as viewed from the character string position, and in Example 4, the present application is continued. Are scanned in the order of cited references.

  FIG. 4 is an operation flowchart showing the control operation of the first embodiment of the example device for the word or word combination in the patent document having the configuration shown in FIG. 2A. Each process in this operation flowchart is executed by a computer system equipped with, for example, a CPU (Central Processing Unit), a memory, and an external storage device.

  First, by searching the patent database 101 (FIG. 2A), a file of a patent document (hereinafter referred to as “displayed patent”) designated by the user is searched (step S401). This process realizes the function of the patent document search unit 102 in FIG. 2A. The patent database 101 is stored on a storage device. FIG. 12A is a diagram illustrating a data configuration example of the patent database 101. As shown in FIG. 12A, the record data of each patent document registered in the patent database 101 includes an application number, a publication number, a summary index, a claim index, an embodiment index, an invention name data, and a patent document. It consists of each field of the body. The index of the embodiment may be an index for implementing the invention. The user expands the patent document file to the memory work area by using keyword data such as application number, publication number, abstract word, claim word, example word, word included in the name of the invention, etc. Patent databases can be searched. When the application number is specified by search, it is searched whether there is a record that matches the character string of the application number specified for search in the application number field in the patent database 101. When the public number is specified for search, it is searched whether there is a record that matches the character string of the public number specified for search in the public number field in the patent database 101. When a word included in the abstract, claim, embodiment, or invention name is specified by search, the search specification is made in the abstract, claim, embodiment index field, or invention name field in the patent database 101. It is searched whether or not the character string of the entered word is registered. As a result, the text data registered in the body field of the retrieved record is extracted.

  Next, in order to search for notice of reasons for refusal, etc., a character string indicating the application number is extracted from the patent document searched in step S401 (step S402). Specifically, the character string registered in the application number field of the record detected in step S401 is extracted. This process realizes the function of the application number extraction unit 103 in FIG.

  Next, the rejection reason database 105 (FIG. 1) is searched for the display patent which is the patent document searched in step S401. The rejection reason database 105 is stored in a file of a storage device such as an external storage device 3005 in FIG. More specifically, a search is made for a file in which a notification of reasons for refusal for a displayed patent and a notification of reasons for refusal quoting the displayed patent are expanded in the work memory (step S403 in FIG. 4). FIG. 12B is a diagram illustrating a data configuration example of the rejection reason database 105. As shown in FIG. 12B, each record of reasons for refusal or the like registered in the reason for rejection database 105 is made up of each field of the text of the citation source application number, citation destination information, and the reason for refusal. When searching for a notice of reason for refusal for the displayed patent, etc., a record in which the application number extracted in step S402 matches the registered content of the citation source application number field of the reason for rejection database 105 is searched. When a notice of reason for refusal that cites the displayed patent is retrieved, a record in which the application number extracted in step S402 is included in the citation destination information field of the reason database 105 is retrieved. As a result, the text data registered in the body field of the retrieved record is extracted. This process realizes the function of the rejection reason notification search unit 104 in FIG. 2A. In this case, the rejection reason database includes a document indicating the reason for refusal, for example, a rejection decision, a trial decision, a pre-examination report, etc., in addition to a notification of the reason for refusal. In the following description, the “text indicating the reason for refusal” combined with these will be referred to as “notification of reason for refusal”.

  Next, from the notification of reasons for refusal searched in step S403 and the display patent (patent document) obtained in step S401, the data of the word pair is obtained in accordance with the control operation shown in the operation flowchart of FIG. Extracted (step S404). This process realizes the function of the word pair extraction unit 106 in FIG. 2A.

Next, it is determined whether or not the word pair data has been extracted in step S404 (step S405).
If the determination in step S405 is YES, the character string of the word to which a specific example is added is specified in the character string of the claim in the display patent claim obtained in step S401 ( Step S406). That is, each character string of a word that is paired with each character string of each abstract word included in the character string of each word pair obtained by the word pair extraction process in step S404 is represented on the character string of the claim. Identified. This process realizes the function of the specific example addition target specifying unit 108 in FIG. 2A.

  Then, the specific example addition target words in the claims identified in step S406 are compared with the corresponding specific examples and displayed on the display of the user's computer (step S407). Alternatively, a link may be set as a specific example addition target, and a specific example may be displayed when the user clicks. This process realizes the function of the display unit 109 in FIG. 2A. However, the specific examples described here are exemplifications of words corresponding to the words to be added to the specific examples in the claims, and finally refer to the second embodiment to determine whether the words are specific examples. Will be described later.

  If the word pair data cannot be extracted in step S404 and the determination in step S405 is NO, the word pair is not displayed, and the words in the claims for the current display patent designated by the user are exemplified. The process ends.

  As described above, in step S403, two types of reasons for refusal notification, such as a notice of reason for refusal for the displayed patent and a notice of reason for refusal that cites the indicated patent, are searched for the following reason. That is, in the notice of reasons for refusal of the displayed patent, as shown in FIG. 5A (a), for example, a description format “Correspondent Document 1“ Smallness and Rumor ”Corresponds to“ Individual Evaluation ”in Claim 1” "Claim 1" indicates claim 1 of the displayed patent (patent document searched in step S401). On the other hand, in the notice of reason for refusal that cites the displayed patent, as shown in FIG. 5 (b), for example, “individual evaluation” in cited reference 2 corresponds to “deflection or custom” in claim 2. In the description format, “Cited document 2” indicates the displayed patent (patent document searched in step S401).

  In both description formats, the concrete word “defamatory or rude” and the abstract word “individual evaluation” are separated by the cited document specific pattern “cited document” and the present application specific pattern “claim”. It is extracted from the part. The “present application” here means the “present application” in the notice of reasons for refusal or the like.

  However, in the description format of FIG. 5A (a), the notice of reasons for refusal regarding the display patent including “claim 1” or the like refers to “cited document 1”. The word “individual evaluation” included in the part after “claim 1”, which is the specific pattern of the present application, is an abstract word included in the display patent. Further, the word “defamatory or rumor” included in the previous part from “Claim 1” which is the specific pattern of the present application is a specific word, that is, a specific example.

  On the other hand, in the description format of FIG. 5A (b), conversely, the notice of reasons for refusal regarding other patents including “Claim 2” and the like refer to “Cited document 2” which is a display patent. The word “defamatory or rumor” included in the part after “claim 2” which is the specific pattern of the present application is a specific word, that is, a specific example. Further, the word “individual evaluation” included in the part before “Claim 2” which is the present application specific pattern is an abstract word included in the display patent.

  Thus, both the notice of reasons for refusal for the displayed patent (FIG. 5A (a)) and the notice of reasons for refusal that cite the indicated patent (FIG. 5A (b)) are included in the notice of reasons for refusal. Based on the cited document specific pattern and the present application specific pattern, word pairs are extracted for the next processing.

  However, the notice of reason for refusal for the displayed patent (FIG. 5A (a)) and the notice of reason for refusal etc. (FIG. 5A (b)) citing the indicated patent are the abstract words in the displayed patent. The positional relationship of the parts for extracting specific words in the notice of reasons for refusal is reversed.

  As described above, the abstract in the display patent can be obtained from either the notification of the reason for refusal regarding the display patent (FIG. 5A (a)) or the notification of the reason for refusal quoting the display patent (FIG. 5A (b)). The correspondence between specific words and specific words in the notice of reasons for refusal, that is, word pairs can be extracted. For this reason, in step S403, the rejection reason notifications in both FIG. 5A (a) and FIG. 5A (b) are retrieved from the rejection reason database 105 (FIG. 2A).

  Next, an outline procedure of the word pair extraction processing in step S404 for extracting word pairs in the first embodiment will be described. About the present application while the patterns from Example 1 to Example 4 described above are discriminated by pattern matching processing, position search processing, and unnecessary character deletion processing by a computer based on the following procedures 1, 2, 3-1, or 3-2. And the description part of the cited document are discriminated and divided, and a pair of a word used in the claim and a specific example word in the cited document is extracted.

Step 1 :
From a document (character string) indicating the reason for refusal retrieved corresponding to a patent document file designated by the user, a sentence having a pair expression pattern character string such as “equivalent” is converted into a plurality of character pattern patterns. Extract by matching process. Here, the pair expression pattern is a character string indicating that the previous word and the subsequent word are semantically the same or paired, as in “equivalent”, and a predetermined one is a dictionary. Stored in A sentence is, for example, a character string starting from the character next to a punctuation and extending to the next punctuation. Regarding the pair expression pattern, the data structure of the table will be described later with reference to FIG.

Step 2 :
Whether one sentence extracted in step 1 has a way of writing the word of the cited document after the word of the present application is written first and the paired expression pattern character string “equivalent” (exception rule → example 4 in FIG. 3D) ) Or not (normal processing → example 1 in FIG. 3A, example 2 in FIG. 3B, example 3 in FIG. 3C).

Procedure 3-1: (Normal Processing → Example 1 in FIG. 3A, Example 2 in FIG. 3B, Example 3 in FIG. 3C):
The description part of one sentence extracted in the procedure 1 is divided by the cited document specific pattern such as “cited document” and the present application specific pattern such as “this application”.

The text before the application specific pattern is extracted as a character string of “citation document”, and the text after the application specific pattern is extracted as a character string of “word of this application”.
In the character string of each extracted part, the unnecessary character string is deleted.

  Regarding the application specific pattern, cited reference specific pattern, and unnecessary character string, the data structure of the table will be described later with reference to FIG.

Procedure 3-2: (Exception rule → Example 4 in FIG. 3D):
The description portion of the single sentence character string extracted in the procedure 1 is divided by a pair expression pattern character string, for example, a character string of “equivalent”. That is, the character string of the pair expression pattern character string is extracted as a “word of reference document” character string, and the character string before the counter expression pattern character string is extracted as a “word of the present application” character string.
Delete unnecessary character strings in the data of each extracted part.

  Next, the details of the word pair extraction process (the function of the word pair extraction unit 106 in FIG. 2A) in step S404 in FIG. 4 will be described according to the operation flowchart shown in FIG. 5B. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

  As described above in step S403 of FIG. 4, when word pair data is extracted from the text data of the displayed patent and the text data of the reason document for rejection corresponding thereto, the cited document specific pattern and the present application specific pattern are The extraction process is executed while being determined. The patterns from Example 1 to Example 4 shown in (a) to (d) of FIG. 3 are discriminated by the pattern matching process and the position detection process based on the above-described procedures 1, 2, 3-1, or 3-2. However, the description part of the present application and the description part of the cited document are discriminated and divided. Hereinafter, the process of the operation flowchart of FIG. 5B will be described with reference to the patterns from Example 1 to Example 4 shown in (a) to (d) of FIG. 3.

  First, it is determined by pattern matching processing whether or not a pair expression pattern character string, for example, a character string of “equivalent” exists in text data such as a rejection reason notice (step S501). As a specific process of step S501, a pattern matching process for searching for a matching of a pair expression pattern character string is executed while reading one character at a time from the top of the rejection reason database. Now, in the expression pattern dictionary 107 of FIG. 2A, for example, a counter expression pattern table as illustrated in FIG. 6A is stored. In this counter-expression pattern table, counter-expression patterns such as “equivalent”, “correspondence”, “equivalent”, “equal”, etc., that link the words in the patent of the present application and the words in the cited document in the notice of reasons for refusal are registered Yes. In step S501, it is determined whether or not a character string such as “equivalent” registered in the counter expression pattern table exists in the file of the rejection reason notice or the like. The expression pattern dictionary 107 is stored in a storage device such as an external storage device 3005 in FIG.

If there is no pair expression pattern character string data and the determination in step S501 is NO, the word pair data is not extracted and the processing of the operation flowchart of FIG. 5B is terminated.
If the determination in step S501 is YES, first, as procedure 1 described above, in the text data of the reason for refusal, for example, a single sentence character string in which the counter-expression pattern character string “equivalent” exists is extracted, and the character string Y and (Step S502). This character string Y is held on a work memory (for example, corresponding to a memory 3002 in FIG. 27 described later). For example, in the description pattern of Example 1 shown in FIG. 3A described above, one sentence shown in FIG. 7A is extracted. Further, in the description pattern of Example 2 shown in FIG. 3B, one sentence data shown in FIG. 8A is extracted. Further, in the description pattern of Example 1 shown in FIG. 3C, one sentence data shown in FIG. 9A is extracted. Further, in the exception description pattern of Example 4 shown in FIG. 3D, one sentence data shown in FIG. 10A is extracted.

  Next, as the above-mentioned procedure 2, in the character string Y, the word of the present application is first written as the first character string following the character string pattern “to”, and the counter-expression pattern character string “corresponds”. It is determined whether there is a description pattern in which the word of the cited document is written as the second character string after the character string (step S503). This is a process of determining whether or not the exception rule pattern corresponds to Example 4 in FIG. Specifically, step S502 is a process of determining whether or not there is a description pattern such as “xxx corresponds to ...” while the character string Y on the work memory is scanned. . More specifically, first, the pattern “to” is first searched by pattern matching while reading one character at a time from the beginning of the character string Y on the work memory. If the pattern matches, the pattern “corresponding to” is searched after “in” of the character string Y. For example, in the description patterns of Example 1, Example 2, and Example 3 shown in FIGS. 3A, 3B, or 3C, the determination in step S503 is NO. In the exception description pattern of Example 4 shown in FIG. 3D, the determination in step S503 is YES.

  If there is no description pattern described above and the determination in step S503 is NO, a series of processing from steps S504 to S512 is executed as data processing corresponding to procedure 3-1.

  First, it is determined whether or not there is a cited document specific pattern in the character string Y (step S504). Here, in the expression pattern dictionary 107 of FIG. 2A described above, for example, a cited document specific pattern table illustrated in FIG. 6B is stored. In step S504, the character string Y of the cited document specific pattern registered in the cited document specific pattern table illustrated in FIG. 6B in the expression pattern dictionary 107 is included while the character string Y on the work memory is scanned. It is determined whether or not. Specifically, each pattern matching process using each cited document specific pattern in the cited document specific pattern table as a search pattern is performed by pattern matching while reading one character at a time from the beginning of the character string Y. Here, every time a cited document specific pattern is detected from the character string Y, a pattern matching process for further detecting the cited document specific pattern in the area of the character string Y after the end of the cited document specific pattern is repeatedly executed. Thereby, as shown in FIG. 3B, a plurality of cited document specific patterns included in the character string Y can be detected.

If the character string Y does not include the cited document specific pattern and the determination in step S504 is NO, the process of the operation flowchart in FIG. 5B is terminated.
If the determination in step S504 is YES, the following pattern matching process is executed while the character string Y on the work memory is scanned in the reverse order of the storage address. That is, when there is no pair expression pattern character string such as “equivalent” or a pair expression pattern character string such as “equivalent” from the back of the searched cited document specific pattern that appears most recently, the end of the character string Y is Extracted as a character string Z. This character string Z is also held on the work memory. Then, while the character string Y on the work memory is scanned, the character string that matches the character string Z on the work memory and the character string that matches the searched cited document specific pattern are deleted. As a result, a new character string Y is obtained on the work memory (step S505). For example, in the description pattern of Example 1 shown in FIG. 3A, “ cited document 2 states that the arbitrating terminal detects a change in the network topology (corresponding to“ communication state information ”in the present application ) in the ad hoc network; . . "(FIG. 7 (a)), the portion" 2 "shown in FIG. 7 (b) shows that the arbitration terminal in the ad hoc network has a network topology (corresponding to" communication status information "in this application ) as a character string Z. In addition, the new character string Y becomes “... In the description pattern of Example 2 shown in FIG. 3B, the portion shown in FIG. Further, in the new character string Y, “... bulletin board” described in cited document 1 corresponds to “content information” of the invention according to claim 1, and “article” described in cited document 1 is claimed in claim 1. This corresponds to the “unit of disclosure of personal opinions” of the invention. " In the description pattern of Example 3 shown in FIG. 3C, the portion shown in FIG. 9B is extracted as the character string Z. Further, the new character string Y becomes “...

  Next, it is determined whether or not the application specific pattern is present in the character string Z on the work memory (step S506). Here, in the expression pattern dictionary 107 of FIG. 6 described above, for example, the application specific pattern table illustrated in FIG. 6C is stored. In step S506, whether the character string of the application specific pattern registered in the application specific pattern table illustrated in FIG. 6C in the expression pattern dictionary 107 is included while the character string Z on the work memory is scanned. It is determined whether or not. Specifically, each pattern matching process using each character string of each application specific pattern registered in the application specific pattern table as a search pattern is performed by pattern matching while reading one character at a time from the beginning of the character string Z. Executed. For example, in the description pattern of Example 1 shown in FIG. 3A, the present application specific pattern of “this application” is included in the character string Z shown in FIG. 7B, so the determination in step S506 is YES. In the description pattern of Example 2 shown in FIG. 3B, the character string Z shown in FIG. 7B includes the application specific pattern “claim”, so the determination in step S506 is YES. In the description pattern of Example 2 shown in FIG. 3C, the character string Z shown in FIG. 9B includes the present application specific pattern “claim of the present application”, so the determination in step S <b> 506 is YES.

  If there is no application specific pattern in the character string Z and the determination in step S506 is NO, the character string of the word pair is not extracted from the character string Z, and steps S504 to S506 are performed for the remaining character string Y. The process is repeated.

  If the determination in step S506 is YES, the character string portion before the character string corresponding to the searched specific pattern in the character string Z on the work memory is extracted as the character string α (step S507). The character string α is stored on the work memory. This character string α corresponds to the words of the cited document. For example, in the description pattern of Example 1 shown in FIG. 3A, from the character string Z in FIG. 7B, the arbitration terminal in the ad hoc network has a network topology in the portion “2” before the specific pattern “this application”. ("Is extracted as the character string α, that is, the word of the cited document, that is, as shown in Fig. 7 (c). In the description pattern of Example 2 shown in Fig. 3 (b), in Fig. 8 (b) From the character string Z, the portion ““ bad words ”described in 1” in front of the specific pattern “claim” of the present application is extracted as the character string α, that is, the word of the cited document, that is, as shown in FIG. In the description pattern of Example 3 shown in Fig. 3 (c), from the character string Z in Fig. 9 (b), the part "1 client computer, A bookstore, `` Each convenience chain is `` '' It is extracted as a leaf, that is, as shown in FIG.

  Subsequently, in the character string Z on the work memory, the character string portion after the character string corresponding to the specific pattern of the present application is extracted as the character string β (step S508). The character string β is also stored on the work memory. In FIG. 5B, “character string β” is displayed as “character string β”. This character string β corresponds to the word of the present application. For example, in the description pattern of Example 1 shown in FIG. 3A, the character string Z in FIG. It is extracted as the column β, that is, the word of the present application. That is, it is as shown in FIG. In the description pattern of Example 2 shown in FIG. 3B, the character string Z in FIG. 8B corresponds to the “individual evaluation” of the portion “1” of the invention according to the specific pattern “claim” of the present application. "Is extracted as the character string β, that is, the word of the present application. That is, it is as shown in FIG. In the description pattern of Example 3 shown in FIG. 3C, from the character string Z in FIG. 9B, the part “1st first information processing device, 3rd” after the specific pattern “this claim” of the present application. Is equivalent to the second information processing apparatus ”is extracted as the character string β, that is, the word of the present application. That is, it is as shown in FIG.

  Next, the character string α (words of the cited document) and the character string β (words of the present application) on the work memory are divided by brackets and punctuation marks (step S509). This is a process for dividing the parallel description of words. Specifically, in each of the character string α and the character string β, each pattern matching process is performed by using the brackets and the punctuation marks as the respective search patterns while reading one character at a time from the beginning, and the key on each character string is executed. Parenthesis and punctuation addresses are detected. Each character string is divided using this address as a delimiter. For example, in the description pattern of Example 1 shown in FIG. 3A, the character string α (words of the cited reference) “2” in FIG. 7C is “a network topology l in the ad hoc network” and the character string β. From (corresponding to the communication state) of (the word of the present application), the character string α (word of the cited document) and the character string β (word of the present application) of FIG. In the description pattern of Example 2 shown in FIG. 3B, the character string α in FIG. 8D is obtained from the character string α (word of the cited document) and the character string β (word in the present application) in FIG. (Word of cited document) and character string β (word of the present application) are obtained. In the description pattern of Example 3 shown in FIG. 3C, the character string α in FIG. 9D is obtained from the character string α (word of the cited document) and the character string β (word in the present application) in FIG. (Word of cited document) and character string β (word of the present application) are obtained. In particular, in the character string α (words of the cited document) and the character string β (words of the present application) in FIG. 9C, the words are described in parallel. For this reason, these words are divided as shown as a character string α (words of the cited document) and a character string β (words of the present application) in FIG.

  Next, the character string corresponding to the unnecessary character string is deleted in each of the character string α (word of the cited document) and the character string β (word of the present application) on the work memory (step S510). Here, in the above-described expression pattern dictionary 107 in FIG. 2A, for example, an unnecessary character string table illustrated in FIG. 6D is stored. In step S510, an unnecessary character string corresponding to each registered pattern in the unnecessary character string table illustrated in FIG. 6D is searched by pattern matching while each of the character string α and the character string β on the work memory is scanned. And deleted. In step S510, dependency analysis is performed on each of the character string α and the character string β on the work memory, and unnecessary character strings are deleted by a dependency analysis process described later.

  In step S510, in each of the character string α and the character string β on the work memory, for example, a clause not related to the last compound word is also deleted as an unnecessary character string. In the description pattern of Example 1 shown in FIG. 3A, the character string “adjustment terminal in ad hoc network” after the unnecessary character string is deleted based on the unnecessary character string table from the character string α in FIG. Consider “network topology”. From this character string, the phrase “the arbitration terminal is in an ad hoc network” not related to the last compound word “network topology” is deleted. As a result, the character string α shown in FIG.

  As a result, the character string α “network topology” shown in FIG. Further, from the character string β “corresponding to the communication state information of the character string β indicating the word of the present application in FIG. 7D, the first hiragana character“ no ”and the symbol“ “”, the character string “equivalent” and the last hiragana 1 The characters “ni” and symbol “” ”are deleted. As a result, the character string β “communication state information” in FIG. In the description pattern of Example 2 shown in FIG. 3B, from the character string α in FIG. 8D, the first numeral “1”, one hiragana character “ni”, “description”, one hiragana character “ ", The symbol" "" "", and the last hiragana character "ga" are deleted. As a result, the character string α shown in FIG. Also, from the character string β in FIG. 8D, the first digit “1” and one hiragana character “ni”, the character string “related” “invention”, one hiragana character “no”, the symbol ““ ”, the character The column “equivalent”, the single hiragana character “ni”, and the symbol “” ”are deleted. As a result, the character string β shown in FIG. In the description pattern of Example 3 shown in FIG. 3C, the first numeral “1” and one hiragana character “NO” are deleted from “1 client computer” in the character string α of FIG. 9D. . Also, the symbol “A” is deleted from “A bookstore” in the character string α. Furthermore, the character string “each” and the last hiragana character “ga” are deleted from “the convenience chain is each” in the character string α. As a result, the character string α shown in FIG. 9E is obtained. In addition, the first numeral “1” and one hiragana character “NO” are deleted from “1 first information processing apparatus” in the character string β in FIG. 9D. In addition, the character string “equivalent” and the last hiragana character “ni” are deleted from “equivalent to the second information processing apparatus” in the character string β. As a result, the character string β shown in FIG. 9E is obtained.

  Finally, the word obtained from the character string α (word of the cited document) on the work memory and the word obtained from the character string β (word of the present application) on the work memory are associated with each other as described above. It is stored on the work memory as a character string indicating a word pair (step S511). For example, in the description pattern of Example 1 shown in FIG. 3A, the word “network topology” in the character string α in FIG. 7E is associated with the word “communication state information” in the character string β. In the description pattern of Example 2 shown in FIG. 3B, the word “bad words” in the character string α and the word “individual evaluation” in the character string β in FIG. In the description pattern of Example 3 shown in FIG. 3C, the word “client computer” in the character string α in FIG. 9E is associated with the word “first information processing device” in the character string β. . Further, the word “bookstore” in the character string α in FIG. 9E is associated with the word “third information processing device” in the character string β. Furthermore, the word “convenience chain” in the character string α in FIG. 9E is associated with the word “second information processing device” in the character string β.

  When the process for the character string Z on the work memory is completed by the series of processes from step S507 to S511, it is determined whether or not the remaining character string Y on the work memory is empty (step S512). . If the determination in step S512 is NO, the process returns to step S504 to determine whether there is a character string of the cited document specific pattern while the character string Y on the work memory is scanned again. For example, in the description pattern of Example 1 shown in FIG. 3A, the remaining character string Y is detected as “... Therefore, the determination in step S512 is NO. However, since the character string Y does not include the cited document specifying pattern, the determination in step S504 is NO, and the process of the operation flowchart in FIG. Further, in the description pattern of Example 3 shown in FIG. 3C, the remaining character string Y is “...” By the process of step S505. No. However, since the character string Y does not include the cited document specifying pattern, the determination in step S504 is NO, and the process of the operation flowchart in FIG.

  On the other hand, in the description pattern of Example 2 shown in FIG. 3B, the “character board” described in the cited reference 1 is the remaining character string Y in the process of step S505. It corresponds to “content information”, and “article” described in cited document 1 corresponds to “disclosure unit of personal opinion” of the invention according to claim 1. " That is, the character string Y includes a cited document specifying pattern. For this reason, after the determination in step S512 is NO, the determination in step S504 is YES.

  As a result, in step S505 newly executed, “an article described in 1 corresponds to“ an individual opinion disclosure unit ”of the invention according to claim 1” is extracted as a character string Z. Then, the new character string Y becomes "..." Bulletin board "described in cited document 1 corresponds to" content information "of the invention according to claim 1." Further, for the character string Z, the determination in step S506 is YES, and a series of processing from step S507 to S511 is executed, and word pair data is extracted.

  Further, after the determination in step S512 is NO and the determination in step S504 is YES again, in step S505 that is newly executed, “the“ bulletin board ”described in 1” is the “content information” of the invention according to claim 1. "Corresponding to" is extracted as the character string Z. Then, the new character string Y becomes “... Furthermore, for the character string Z, the determination in step S506 is YES, and a series of processing from step S507 to S511 is performed, and word pairs are extracted.

  Finally, the determination in step S512 is NO, the determination in step S504 is NO for the remaining character string Y “..., Do”, and the processing of the operation flowchart in FIG.

  Next, in step S503 of FIG. 5B, the description pattern “xxx” includes the word of the present application written in the character string Y and the word of the cited document after the data “corresponding” of the pair expression pattern character string. When it is determined that there is a “...” corresponding to..., A series of processing from step S513 to S517 is executed. This is data processing corresponding to the above-described procedure 3-2 corresponding to the example 4 in FIG.

  First, while the character string Y on the work memory is scanned, the character string portion after the data of the pair expression pattern character string is extracted as the character string α (step S513). This character string α is stored on the work memory. This character string α corresponds to the words of the cited document. For example, in the exception description pattern of Example 4 shown in FIG. 3D, from the character string Y in FIG. 10B, the part “corresponding” of the pair expression pattern character string ““ movement instruction ”( Scheduling based on paragraph [0100]) is also disclosed. "Is extracted as the character string α, that is, the word of the cited document. That is, it is as shown in FIG.

  Subsequently, in the character string Y on the work memory, the character string portion before the character string data corresponding to the pair expression pattern character string “corresponds to” is extracted as the character string β (step S514). This character string β is also stored on the work memory. This character string β corresponds to the word of the present application. For example, in the exception description pattern of Example 4 shown in FIG. 3D, the meta level task is also changed from the character string Y in FIG. "" Is extracted as the character string β, that is, the word of the cited document. That is, it is as shown in FIG.

  Next, the character string α (words of the cited document) and the character string β (words of the present application) on the work memory are divided by brackets and punctuation marks (step S515). This is a process for dividing the parallel description of words, similar to the process of step S509 described above. For example, in the exception description pattern of Example 4 shown in FIG. 3D, the character of FIG. 10D is obtained from the character string α (word of the cited document) and the character string β (word of the present application) of FIG. A sequence α (words of the cited document) and a character string β (words of the present application) are obtained.

  Next, while the character string α (words of the cited document) and the character string β (words of the present application) on the work memory are scanned, the unnecessary character illustrated in FIG. Unnecessary character strings corresponding to each registered pattern in the character string table are searched and deleted, or unnecessary character strings are deleted by dependency analysis (step S516). For example, in the exception description pattern of Example 4 shown in FIG. 3D, the character string α “movement instruction” in FIG. can get. Further, the character string β “meta-level task” in FIG. 10E is obtained from the “meta-level task” in the character string β “1” in FIG.

  Finally, the word obtained from the character string α (word of the cited document) on the work memory and the word obtained from the character string β (word of the present application) on the work memory are associated with each other as described above. It is stored as data indicating word pairs on the work memory (step S517). For example, in the exception description pattern of Example 4 shown in FIG. 3D, the word “move instruction” in the character string α in FIG. 10E and the word “meta-level task” in the character string β are associated with each other.

In the processing of the first embodiment described above, words obtained from the character string α (words of the cited document) and words obtained from the character string β (words of the present application) in step S511 or S517 in FIG. Corresponding and stored as data indicating word pairs on the work memory. In response to this, each data portion of the word that is semantically the same as or paired with each abstract word data included in the word pair data is added to the claim data in the claims of the displayed patent. Specified by
Through the series of processes from step S513 to S517, the process for the character string Y on the work memory is completed, and the process of the operation flowchart of FIG. 5B is completed.

  FIG. 11 shows a display example of the display unit. Along with the claim 1101 of the displayed patent, the correspondence 1102 between the words described in the claims, for example, “content information” and the words described in the cited document, for example, “bulletin board” is displayed.

  FIG. 13A is an operation flowchart showing specific operations of the pattern matching process and the character string position detection process in the process of steps S501, S503, S504, S506, S509, S510, S513, S515, S518, or S519 of FIG. is there. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

  In FIG. 13A, first, pattern character string data to be searched is read from a work memory or a storage device into a predetermined area on the work memory (step S601). For example, in step S501 of FIG. 5B, the pattern character string is a character of each pair expression pattern read from the pair expression pattern table shown in FIG. 7A in the expression pattern dictionary 107 (FIG. 2A) on the storage device. Is a column.

  Next, data of a matching target character string, which is a character string to be subjected to pattern matching processing, is read from a work memory or a storage device into a predetermined area on the work memory (step S602). For example, in step S501 of FIG. 5B, the matching target character string is a file such as a rejection reason notice read from the rejection reason database 105 on the storage device.

  Next, the address of the first character of the pattern character string read in the predetermined area on the work memory in step S601 is set to the pattern pointer p1 held in the variable area on the work memory (step 603).

  Next, the address of the first character of the character string to be matched read in the predetermined area in the work memory in step S602 is set in the matching base point pointer p2 and the matching pointer p3 held in the variable area on the work memory ( Steps S604 and S605).

  Next, while the address of the matching pointer p3 is incremented by 1 (step S607), a character that matches the first character in the pattern character string indicated by the pattern pointer p1 is searched from the matching target character string (step S606). (Repeat processing of S607 → S608 → S606).

  As a result of incrementing the matching pointer p3, if the value of p3 exceeds the end address of the character string to be matched read into a predetermined area on the work memory, no matching is output and the processing in FIG. 13A is performed. Is finished (steps S608 → S609).

  When the first character in the pattern character string indicated by the pattern pointer p1 matches in the character string to be matched and the determination in step S606 becomes YES, the following series of processing from step S610 to S617 is executed. Here, it is determined whether or not the character string continuing from the matching position in the matching target character string matches all the second and subsequent characters of the pattern character string.

First, the address of the matching pointer p3 when a match is detected in step S606 is set to the matching base point pointer p2 on the work memory (step S610).
Next, the address value of the pattern pointer p1 on the work memory is incremented by 1 (step S611).

  Next, it is determined whether or not the address value of the pattern pointer p1 has exceeded the end address of the pattern character string read into the predetermined area on the work memory (step S612).

If the determination in step S612 is NO, the address value of the matching pointer p3 is incremented by 1 (step S613).
Next, it is determined whether or not the address value of the matching pointer p3 exceeds the end address of the matching target character string read into the predetermined area on the work memory (step S615).

  If the determination in step S614 is NO, it is determined whether or not the character at the address position in the matching target character string pointed to by the matching pointer p3 matches the character at the address position in the pattern character string pointed to by the pattern pointer p1. (Step S617). That is, when step S617 is first executed after YES in step S606, the second character counted from the position where the matching position is detected in step S606 in the matching target character string is the two characters of the pattern character string. It is determined whether or not the eye matches.

  If the determination in step S617 is YES, the process returns to step S611 and the series of processes up to step S617 is executed again. Here, the address values of the pattern pointer p1 and the matching pointer p3 are incremented by 1 (steps S611 and S614). As a result, it is further determined whether or not the third character counted from the position where the matching position is detected in step S606 in the matching target character string matches the third character of the pattern character string.

  By repeating the above processing, it is determined whether or not the character string that continues from the position where a match is detected in step S606 in the matching target character string matches all the second and subsequent characters of the pattern character string.

  In the above series of processing, if the value of the pattern pointer p1 incremented in step S611 exceeds the end address of the pattern character string read into the predetermined area on the work memory, the pattern character on the matching target character string A match with the column is detected. In this case, the address value of the matching base point pointer p2 indicating the position where the match is detected in step S606 in the matching target character string is output as the matching result, and the processing in FIG. 13A is terminated (steps S612 → S613).

  On the other hand, if the value of the matching pointer p3 incremented in step S614 exceeds the end address of the matching target character string read into the predetermined area on the work memory, no matching is output and the processing in FIG. 13A is performed. Is finished (step S615 → S616).

  In the above-described series of processing, the character at the address position in the matching target character string pointed to by the matching pointer p3 does not match the character at the address position in the pattern character string pointed to by the pattern pointer p1, and the determination in step S617 is NO. If it becomes, re-search is executed. That is, the address of the first character of the pattern character string read into the predetermined area on the work memory in step S601 is set in the pattern pointer p1 (step 618). Further, the address value corresponding to the position next to the position where the match is detected in step S606 indicated by the matching base point pointer p2 is set in the matching pointer p3 (step S619). Thereafter, the processing returns to step S606, and the following processing is executed while the address of the matching pointer p3 is incremented by 1 from the position next to the matching position (step S607). That is, the character that matches the first character in the pattern character string indicated by the pattern pointer p1 is searched again from the character string to be matched (steps S606 → S607 → S608 → S606 repetitive processing).

  When the first character in the pattern character string indicated by the pattern pointer p1 matches again in the matching target character string and the determination in step S606 is YES, the following series of processing from step S610 to S617 is executed. Thereby, it is determined again whether or not the character string continuing from the matching position in the matching target character string matches all the second and subsequent characters of the pattern character string.

  By repeating the above-described series of processing, a pattern character string is searched for in the matching target character string. If matching is successful, the head position of the match is output as a matching result (step S613).

  Next, unnecessary character deletion processing by dependency analysis in three steps is performed in the flowchart of FIG. 13B and the morpheme dictionary (a), part-of-speech connection probability (b), phrase connection rule (c), phrase connection probability ( This will be described with reference to each table of d). Each process in the operation flowchart is executed by a computer system including, for example, a CPU, a memory, and an external storage device.

Step S1301: First, morphological analysis is executed. The morpheme string having the highest probability is selected using a dictionary (a) of morphemes whose elements are parts of speech and character strings such as nouns, particles, and verbs, and a table (b) defining connection probabilities of parts of speech. The morpheme dictionary (a) and the table (b) defining the connection probabilities of parts of speech are stored, for example, on a work memory or a storage device. Here, for example, a case where one sentence is “The bulletin board described in the cited document corresponds to the content information of the claimed invention” will be described. Probabilities are calculated for the following morpheme sequence candidates, and the morpheme sequence with the highest probability is selected.
(1) Citation (sa variable noun), literature (noun), (particle), description (sa variable noun), (particle). . .
(2) Citation (sa variable noun), sentence (noun), dedication (unregistered word), ni (particle), description (sa variable noun). . .

Step S1302: Next, phrase division processing is performed. That is, a phrase is generated by connecting morphemes according to the phrase connection rule (c).
In the above example, the following clause is generated.
“In cited document”, “Described”, etc.

Step S1303: Next, dependency analysis is performed. That is,
(1) The dependency destination with the highest probability is selected according to the connection probability (d) of the clause and the constraint rule.
(2) The connection probability (d) is defined by the phrase type and proximity relationship.
For example, in the above example, the types of clauses are as follows: in the cited document, a noun clause with a “Ni” case, and with a description, a noun clause with a “no” case.
(3) Perform proximity processing. For example, it is calculated as follows, “in cited document” and “description” are adjacent, and “in cited document” and “bulletin board” are one-segment intervals.

On the other hand, a dependency destination that satisfies the following constraints and has the maximum probability is obtained.

・ The clauses other than the last one always have one dependency clause behind. ・ The dependency relationship does not intersect.

Finally, the following dependency pair is generated from the above example.

・ "In the cited document-described", "described in-bulletin board", "bulletin board-equivalent", etc.

  In the dependency analysis process described above, the dependency pair is generated by pattern matching with the dictionary and calculation of the probability attached to the matched pattern. Note that the above-described dependency analysis may be executed so that the entire dependency relationship is extracted using the original character string Y, not only from the character string α or β.

  Even if the inventor has not clearly shown an example in the displayed patent, the first embodiment of the example of the word or the combination of words in the patent document described above can provide a deep understanding of the contents of the displayed patent. The interpretation of the examiner who prepared the notice of reasons for refusal corresponding to the indicated patent is used. Thereby, it is possible to know what the words in the claims specifically represent, and even a user who is not familiar with the claim description method can understand the claims.

  In the above description, specific examples have been described as being used in the words used in the claims. However, by applying the same processing to words other than the claims, for example, the embodiment Specific examples can also be displayed in the words used in the embodiments for carrying out the invention.

  FIG. 14 is a configuration diagram of a second embodiment of an example device for words or combinations of words in a patent document. In FIG. 14, the same numbers are assigned to parts that perform the same processing as in the first embodiment shown in FIG. 1.

  The configuration of the second embodiment includes the following configuration in addition to the configuration of the first embodiment shown in FIG. That is, a specific example determination unit 1401 that determines which one of the two words constituting each word pair extracted by the word pair extraction unit 106 is a specific example by referring to the word pair database 1402 is included. . The specific example determination unit 1401 further includes a patent document analysis unit 1401-1, a word appearance position specifying unit 1401-2, a patent document search hit number acquisition unit 1401-3, and a determination unit 1401-4.

In the first embodiment, it may not be possible to simply determine which of the two words is a specific example from only a pair of words.
If you use the description of the reason for refusal without dropping the information to the pair of words, the information of the section in which the word is used (the section in the description, such as claims, examples or forms for carrying out the invention) May be used to estimate which is specific, but the section information is not always written.

For example, in the following notice of reasons for refusal, the words of cited documents are more specific.
Example of notification of reasons for refusal 1:
“The“ bulletin board ”described in the cited document 1 corresponds to the“ content information ”of the invention according to claim 1, and the“ article ”described in the cited document 1 is the“ disclosure unit of the individual opinion ”of the invention according to claim 1. The “bad word” described in the cited document 1 corresponds to “individual evaluation” of the invention according to claim 1. ]
Example 2 of notification of reasons for refusal:
“In Cited Document 2, an arbitration terminal detects a change in network topology (corresponding to“ communication status information ”in the present application) in an ad hoc network.”

On the other hand, in the notification of reasons for refusal as described below, the words of the present application (or the claimed invention) are more specific.
Example 3 of notification of reasons for refusal:
“The“ signal processing circuit 14 ”of the invention described in the cited document 1 corresponds to the“ route selection means ”of the invention according to claim 1,”
Example 4 of notification of reasons for refusal:
“The“ plate spring-like member ”described in the cited document 2 can be said to support the rear part of the fish body from the inside. Therefore, the“ plate spring-like member ”is the“ rib-like shape ”according to claim 2 of the present application. It can be said that it corresponds to “part”. ]

  Here, in general, a word used in a claim is more likely to be abstract than a word used in an embodiment or a mode for carrying out the invention (hereinafter referred to as “example etc.”). . Therefore, in the second embodiment, the specific example determination unit 1401 analyzes information on which section each word is used for each word constituting the word pair extracted by the word pair extraction unit 106. Based on the analysis result, it is determined which of the words constituting the word pair is a specific example.

  More specifically, the specific example determination unit 1401 executes a determination process in which the following two methods are combined.

Method 1: In the patent database 101 (see FIG. 12A), with respect to the patent document of the cited document in the notice of reasons for refusal and the patent document of the present application, the appearance position (section) where each word constituting the extracted word pair is used A determination process is performed by specifying (). That is, it is determined whether the appearance position (section) is a section of [Claims] or other sections such as [Embodiment of the Invention], [Summary of Invention], [Summary], and the like.

Method 2: When there is a word used in both the claim of any patent document and the mode for carrying out the invention in Method 1, the patent database 101 ( The entire patent document set registered in FIG. 12A) is searched. Then, determination processing is performed based on the number of hits as a result of each search.

  FIG. 15 is an operation flowchart showing the control operation of the second embodiment of the example device for words or word combinations in the patent document having the configuration shown in FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device. In FIG. 15, the same numbers S401, S402, S403, S405, and S406 are assigned to parts that perform the same processing as in the first embodiment shown in FIG. 4. In the control operation of the second embodiment, the following control operation is executed in addition to the control operation of the first embodiment. That is, for each word pair data for which the word pair data can be extracted in step S404 in FIG. 15 and the determination in step S405 is YES, which of the two words constituting each word pair is a specific example. The process which determines whether is performed (step S1501). The processing in step S1501 realizes the function of the specific example determination unit 1401-4 in FIG.

  FIG. 16 is an operation flowchart showing a specific control operation in step S1501 of FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device. The processing here is repeatedly executed for each pair of words extracted in step S404 in FIG. The specific example obtained as a result is registered in a word pair table to be described later.

  First, in the determination process according to the above-described method 1, that is, in each text of the patent document of the cited document and the patent document of the present application in the notice of reasons for refusal, the appearance position where each word constituting the pair of extracted words is used ( A determination process is performed by specifying (section) (step S1601). This process realizes the functions of the patent document analysis unit 1401-1 and the word appearance position specifying unit 1401-2 in FIG.

Next, it is determined whether or not the determination has been made by the method 1 (step S1602). This process realizes a part of the function of the determination unit 1401-4 in FIG.
If the determination can be made by the method 1, the process of the operation flowchart of FIG. On the other hand, when the determination cannot be made by the method 1, the determination process by the method 2 described above is executed (step S1603). That is, the entire predetermined patent document set registered in the patent database 101 (see FIG. 12A) is searched by each word constituting the extracted word pair. Then, determination processing is performed based on the number of hits as a result of each search. This process realizes the function of the patent document search hit number acquisition unit 1401-3 and the partial function of the determination unit 1401-4 in FIG.

  FIG. 17 is an operation flowchart showing the detailed control operation of method 1 in step S1601 of FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

  First, the file of the cited document patent document and the file of the patent document of the present application specified in the file such as the notice of reason for rejection searched in step S403 of FIG. 15 are acquired from the patent database 101 (step S1701). Each patent document file is developed on the work memory.

  Next, the section structure is analyzed in each patent document file acquired on the work memory in step S1701 (step S1702). Specifically, by searching for a character string corresponding to each heading enclosed in black bracket tags in each patent document file, a section of each claim in the claims and a description A section in which an embodiment including the example is described is analyzed.

  Next, in each patent document acquired in step S1701, the character string of word A and the character string of word B constituting the word pair extracted in the word pair extraction process (step S404 in FIG. 15). Is identified (section) (step S1703). At this time, the analysis result of the section in step S1702 is referred to. The word A is a word extracted in association with the cited document patent document designated in the notice of reasons for refusal or the like. The word B is a word extracted in association with the patent document of the present application designated in the notice of reasons for refusal or the like. The word pair data extracted in step 404 is registered in the word pair database 1402 having the data configuration example shown in FIG. The word pair database 1402 is stored in a storage device. Each record of the word-pair database 1402 includes citation information, that is, information specifying a cited document (publication number, document name, etc.) as information about a cited document cited as a cited document in a cited party, that is, a notice of reason for refusal. Has a field for registering. Each record has a field for registering the word at the quotation destination extracted in step S404 in association with the quotation destination, that is, the cited document. Further, each record has a field for registering citation source information, that is, an application number for identifying the present application, as information relating to a patent document cited as the present application in the citation, that is, a notice of reasons for refusal. Each record has a field for registering a word at the quotation destination extracted in step S404 in association with the quotation source, that is, the present application. In addition, each record has a specific example determination result field. The contents of this field are blank when the word pairs extracted in step S404 are registered in each record. As a result of the execution of the specific example determination process in step S1501, the word determined as a specific example among the words registered in the word field at the quotation destination and the words registered in the word field at the quotation source is determined. be registered.

  Next, an embodiment in which the word A data appears in the text data of the cited patent document only in the claims section, and the data of the word B includes examples in the text data of the present patent document Or the like (it may appear in the claim section at the same time) or not (step S1704).

  If the determination in step S1704 is YES, it is determined that word B is a specific example (step S1705). Then, the operation flowchart of FIG. 17, that is, the process of step S1601 of FIG. 16 ends.

  If the determination in step S1705 is NO, the following determination is further performed. That is, the data of the word A appears in the embodiments including the examples in the text data of the cited patent document (may appear in the claims section at the same time), and the data of the word B is the patent document of the present application. It is determined whether or not the place that appears in the text data is only the section of the claim (step S1706).

If the determination in step S1706 is YES, it is determined that word A is a specific example (step S1707). Then, the operation flowchart of FIG. 17, that is, the process of step S1601 of FIG. 16 ends.
If the determination in step S1706 is NO, the determination is not made in Method 1, and the operation flowchart in FIG. 17, that is, the process in step S1601 in FIG.

  When the control operation of FIG. 17 is executed, the word pair data extracted in step 404 in FIG. 4 is registered in the word pair table having the data configuration example shown in FIG. The word pair table is held on the work memory. Each row of the word pair table has a field for registering the word A extracted in step S404 in association with the cited document patent document in the cited side, that is, the reason for rejection. Each line has a field for registering the appearance position (section) of the word A in the cited patent document based on the processing in step S1703. Further, each line has a field for registering the word B extracted in step S404 in association with the patent document of the present application in the citation side, that is, the reason for rejection. In addition, each line has a field for registering the appearance position (section) of the word B in the cited patent document based on the processing in step S1703. Each row has a fold for registering a specific example determination result based on the appearance position. In this field, of the words registered in the word A field and the word B field, words based on the results determined as specific examples in steps S1704 and S1706 based on the contents of the corresponding appearance position fields are registered. The

  For example, in each line indicated by 1801 in FIG. 18, the appearance position of the data of the word A “client computer, bookstore, convenience chain” in the cited example (cited document) is an example in the text data of the cited document. And the determination in step S1704 in FIG. 17 is NO. In addition, only the section of the claim in the text data of the patent document has the appearance position of the data of the citation side (this application) word B “first information processing device, third information processing device, second information processing device”. Appears in step S1706, the determination in step S1706 becomes YES. As a result, in these lines, the words in the word A field are specific examples, so the words registered in the word A field are registered in the registration field of the specific example determination result “client computer, bookstore, convenience chain”. The In the lines other than the respective lines indicated by 1801 in FIG. 18, the appearance position of the data of the word A “movement instruction, bad mouth, leaf spring portion” appears in the section of the example etc. in the text data of the cited document. Therefore, the determination in step S1704 is NO. In addition, since the appearance position of the word B data “metal bell task, personal evaluation, rib-like portion” appears in both the claims and the embodiments in the text data of the patent document of the present application, The determination is also NO. For this reason, in the lines other than the lines indicated by 1801, both the word A and the word B appear in the embodiment, and therefore it cannot be determined which is a specific example. Therefore, in the operation flowchart of FIG. Registration of the determination result in the registration field is not performed.

  FIG. 19 shows the detailed control operation of method 2 in step S1603 in FIG. 16, which cannot be determined by method 1 in step S1601 in FIG. 16, but is determined as NO in step S1602 in FIG. It is a flowchart. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

  First, in the patent document developed on the work memory in step S1701 of FIG. 17, only the section of the claim determined based on the analysis result of the section in step S1702 of FIG. Then, the number of hits CA1 when the search range is searched with the character string corresponding to the word A is calculated (step S1901).

  Next, in the patent document developed on the work memory in step S1701 in FIG. 17, only the section such as the embodiment determined based on the analysis result of the section in step S1702 in FIG. Then, the number of hits CA2 when the search range is searched with the character string corresponding to the word A is calculated (step S1902).

  Subsequently, in the document developed on the work memory in step S1701 in FIG. 17, only the section of the claim determined based on the analysis result of the section in step S1702 in FIG. Then, the number of hits CB1 when the search range is searched with the character string corresponding to the word B is calculated (step S1903).

  Furthermore, in the document file acquired on the work memory in step S1701 in FIG. 17, only the section of the embodiment or the like determined based on the analysis result of the section in step S1702 in FIG. Then, the hit number CB2 when the search range is searched with the character string corresponding to the word B is calculated (step S1904).

  Then, the value of CA2 / CA1 and the value of CB2 / CB1 are calculated, and each calculation result is held in the work memory. CA2 / CA1 is a value indicating how much the word A appears in the section of the embodiment in the patent document with respect to the section of the claims. Similarly, CB2 / CB1 is a value indicating how much the word B appears in the section of the embodiment etc. with respect to the section of the claims in the patent document of the present application. As a result, it is determined whether the value of CB2 / CB1 on the work memory is larger than the value of CA2 / CA1 on the work memory (step S1905).

  If the determination in step S1905 is YES, it is determined that the word B is a specific example (step S1906). In this case, the ratio of the word B appearing in the section of the example etc. in the patent document is larger than the ratio of the word A appearing in the section of the example etc. in the patent document.

On the other hand, if the determination in step S1905 is NO, the word A is determined to be a specific example (step S1907). In this case, the ratio of the word A appearing in the section of the example etc. in the patent document is larger than the ratio of the word B appearing in the section of the example etc. in the patent document.
The above determination is based on an empirical rule that specific words have a higher probability of being used in sections such as examples.

  At the time of execution of the control operation of FIG. 19 described above, the word pair of the line that was not registered in the specific example determination result field of the word pair table illustrated in FIG. 18 on the work memory in the control operation of FIG. For example, the movement instruction and the meta level task are registered in the word pair table having the data configuration example shown in FIG. This word pair table is held in the work memory. Each row of the word pair table registers the word A (movement instruction, bad mouth, leaf spring-like member) extracted in step S404 in association with the cited document patent document in the cited side, that is, the reason for rejection. Has a field. Each row has fields for registering the numbers of hits CA1 and CA2 calculated in steps S1901 and S1902, respectively. Furthermore, each line has a field for registering the word B (meta-level task, personal evaluation, rib-shaped portion) extracted in step S404 in association with the patent document of the present application in the citation side, that is, the reason for rejection. Each row has fields for registering the hit numbers CB1 and CB2 calculated in steps S1903 and S1904, respectively. Each row has a fold for registering a specific example determination result based on the number of hits. In this field, words based on the result determined in step S1905 based on the ratios of CA2 / CA1 and CB2 / CB1 are registered.

  For example, in each row indicated by 2001 and 2002 in FIG. 20, since the ratio of CA2 / CA1 is larger than the ratio of CB2 / CB1, the words registered in the word A field (movement instructions, bad words) It is registered in the registration field of the specific example determination result. Further, in the row indicated by 2003 in FIG. 20, since the ratio of CB2 / CB1 is larger than the ratio of CA2 / CA1, the word registered in the word B (rib-shaped part) field is a specific example determination result. Registered in the registration field.

  As described above, based on the contents of the word pair table illustrated in FIG. 18 obtained by Method 1 and the contents of the word pair table illustrated in FIG. A specific example determination result corresponding to the pair is obtained. The specific example determination results obtained in these word pair tables are registered in the word pair database 1402 (FIG. 14) illustrated in FIG. As a result, the determination process of the specific example in step S1501 in FIG. 15 ends.

  Thereafter, in the same manner as in the first embodiment of FIG. 4, in the claims in the claims in the display patent obtained in step S401 of FIG. Is identified (step S406). That is, in each word pair newly registered in the word pair database 1402 exemplified in FIG. 23, each part of the same word as the word registered in the specific example determination result field (FIG. 23) is Specified by

  Then, the specific example addition target words in the claim specified in step S406 illustrated as 1101 in FIG. 11 are compared with the specific examples corresponding thereto, and are illustrated in 1102 in FIG. Is displayed on the display of the user's computer (step S407).

  FIG. 21 is a diagram illustrating an operation flowchart of another embodiment of the control operation of the determination processing of the specific example of step S1501 of FIG. 15 in the second embodiment. This embodiment is not limited to the present application and cited documents. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

  In this control operation, a set of patent documents in which both of two words are used in one case, instead of obtaining the number of two words constituting the word pair independently from the patent database 101. Is created. Then, which one is more specific is determined by the appearance location of the two words in the set.

  First, a patent document file including both the words A and B constituting the word pair data extracted in the word pair extraction process (step S404 in FIG. 15) is obtained from the patent database 101 (FIG. 14). A patent set obtained as a result of the search is created and stored in the work memory (step S2101). That is, it is as 2101 in FIG.

Next, the value zero is assigned to both the number variables D1 and D2 held in the work memory (step S2102).
Next, the processing from step S2102 to step S2108 is repeatedly executed for each patent document file in the patent set created in step S2101 of FIG.

  First, in the current patent document file, the correspondence relationship between the section of each claim in the claims and the paragraph in the section of the embodiment or the like is analyzed (step S2103). This is because the words A and B are selected from the claims and the corresponding embodiments.

  FIG. 22 is an operation flowchart showing the detailed operation of the analysis processing in step S2103 of FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device.

First, text information is read from the current patent document file (step S2201).
Next, the data of the text document read in step S2201 is divided for each section by a black bracket tag (step S2202).

Next, among the sections divided in step S2202, the text data portion of the document in the claim section is divided for each item in the claim (step S2203).
Next, keyword data is extracted from each item of the claim divided in step S2203 (step S2204). The keyword data is for finding a biller and an example corresponding to it, and can be extracted by executing, for example, a morphological analysis process on the text data of each item of the claim.

Next, the data of the text document in the section of the embodiment or the like read in step S2201 is divided into paragraphs by using a bracket with a black bracket (step S2205).
Next, keywords are extracted from each paragraph part of the embodiment or the like divided in step S2205 (step S2206). The keyword can be extracted by executing, for example, a morphological analysis process on the text data of each paragraph part in the embodiment or the like.

  Finally, for the data corresponding to each item of the claim divided in step S2203, the keyword in step S2205 including the keyword in step S2206 that is most similar to the keyword extracted in step S2204 Data is associated (step S2207).

The process of step S2103 of FIG. 21 is executed as the process of the operation flowchart of FIG.
Returning to the operation flowchart of FIG. 21, after the processing of step S2103, the word A is used in the claims in the current patent document, and the word B is used in the paragraph of the embodiment corresponding to the claims. It is determined whether or not it has been interrupted (step S2104). The paragraphs such as the example corresponding to the claims are selected in step S2103.

If the determination in step S2104 is YES, the number variable D1 on the work memory is incremented, and if NO, the number variable D1 is not incremented (step S2105).
Next, in the current patent document, it is determined whether or not the word B is used in the claim and the word A is used in the embodiment corresponding to the claim (step S2106). In this case as well, paragraphs such as examples corresponding to the claims are selected in step S2103.

If the determination in step S2106 is YES, the number variable D2 is incremented. If NO, the number variable D2 is not incremented (step S2107).
Thereafter, it is determined whether all patent document files in the set created in step S2101 have been processed (step S2108).

If determination of step S2108 is NO, it will return to step S2103 and a process will be repeated.
If the determination in step S2108 is YES, it is finally determined whether or not the number variable D1 is larger than the number variable D2 (step S2109).

  In the paragraphs of claims and examples etc. corresponding to each other, the word A is used on the claims side and the word B is used on the side of the examples etc., the word B is used on the claims side and the word A is executed If the number of cases used on the example side is larger, the word B is recognized as a specific example (step S2010).

  Conversely, the number of cases where the word B is used on the claim side and the word A is used on the side of the embodiment etc. is greater than the number of cases where the word A is used on the claim side and the word B is used on the side of the embodiment. For example, the word A is recognized as a specific example (step S2011).

  When the specific example determination result corresponding to each word pair is obtained by the operation flowchart of FIGS. 21 and 22 as described above, the specific example determination result obtained in the pair table of each word is shown in FIG. It is registered in the illustrated word pair database 1402 (FIG. 14). As a result, the determination process of the specific example in step S1501 in FIG. 15 ends.

Subsequent processing is as described above.
According to the second embodiment described above, it is possible to accurately determine which word corresponds to a specific example by analyzing how two words constituting a word pair appear in a patent document. Can be determined.

  24A and 24B are configuration diagrams of a third embodiment of an example device for words or combinations of words in a patent document. In FIG. 24A and FIG. 24B, the same number is attached | subjected to the part which performs the same process as 2nd Embodiment shown by FIG.

  In the second embodiment shown in FIG. 14, word pairs are extracted by the word pair extraction unit 106 for the patent document designated by the user. Furthermore, specific examples of the words in the claims are exemplified while the specific example determination unit 1401 determines a specific example of the two words constituting the word pair. In this case, word pairs can be stored as a word pair database 1402 illustrated in FIG. 23 together with information indicating which of the two words constituting the word is a specific example.

Therefore, in the third embodiment, the configuration in which the word pair database 1402 is stored in advance and the configuration illustrated while searching for word pairs and specific examples from the word pair database 1402 when a user specifies a patent document are independent. And realized. ,
First, the configuration of FIG. 24A does not include the specific example addition target specifying unit 108 and the display unit 109 as compared to the configuration of FIG. The control operation for realizing the function of this configuration is shown by the operation flowchart in FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device. In FIG. 25, the same number is attached | subjected to the part which performs the same process as 2nd Embodiment shown by FIG.

In the operation flowchart of FIG. 25, a series of processing from step S402 to S1501 is the same as in the case of FIG.
In FIG. 25, after the first patent document is read from the patent database 101 in step S2501, the next patent document is sequentially read from the patent database 101 in step S2504 until it is determined in step S2503 that all patent documents have been processed. Is read. Then, word pair extraction and specific example determination processing are executed for each patent document file by a series of processing from step S402 to S1501, as in the case of the second embodiment. Then, the word pair extraction and the determination result of the specific example are stored in the word pair database 1402 having the configuration illustrated in FIG. 23 in step S2502.

  As a result of the above processing, a pair of words appearing in claims and specific words corresponding thereto can be stored in advance as a word pair database 1402 in the patent document file in the patent database 101.

FIG. 24B is a configuration of the third embodiment that illustrates a patent document designated by the user while searching for word pairs and specific examples from the word pair database 1402.
In FIG. 24B, the patent database 101, the patent document search unit 102, and the application number extraction unit 103 are the same as those in the first embodiment (FIG. 1) or the second embodiment (FIG. 14). That is, when a patent document is designated by the user, the patent document is searched from the patent database 101, and the application number is extracted.

Next, the word pair search unit 2401 extracts words from the claims of the designated patent document, and searches the word pair database 1402 for specific examples corresponding to the words.
The functions of the specific example addition target specifying unit 108 and the display unit 109 are the same as those in the first embodiment (FIG. 1) or the second embodiment (FIG. 14).

  The control operation for realizing the function of the configuration of FIG. 24B is shown by the operation flowchart of FIG. Each process in the operation flowchart is executed by, for example, a computer system including a CPU, a memory, and an external storage device. In FIG. 26, the processing in step S401 and step S402 is the same as that in the first embodiment shown in FIG. 4 or the second embodiment shown in FIG. That is, by searching the patent database 101 (FIG. 1), a patent document (displayed patent) designated by the user is searched (step S401). Next, the application number is extracted from the patent document searched in step S401 (step S402).

  Next, the word pair database 1402 having the data configuration example illustrated in FIG. 23 is accessed, and the citation source information field and the citation destination information field are searched based on the application number extracted in step S402. When the record can be extracted, the words in the specific example determination result field of the record and the words that are paired with the words are extracted (step S2601).

Next, it is determined whether a specific example has been found in step S2601 (step S2602).
The processes in steps S2601 and S2602 described above realize the function of the word pair search unit 2401 in FIG.

  If the determination in step S2602 is YES, the word to which a specific example is to be added is specified in the claims in the claims in the displayed patent obtained in step S401 (step S406). Then, the specific example addition target words in the claim specified in step S406 illustrated as 801 in FIG. 8 are compared with the specific examples corresponding to each, as illustrated in 1102 in FIG. Is displayed on the display of the user's computer (step S407).

  According to the third embodiment described above, it is possible to previously create a word pair database 1402 for illustrating specific examples of words in claims. As a result, when a user designates a patent document, specific examples of words in claims in the patent document can be presented at high speed with reference to the word pair database 1402.

FIG. 27 is a diagram illustrating an example of a hardware configuration of a computer that can realize the system according to each of the above-described embodiments.
A computer shown in FIG. 27 includes a CPU 3001, a memory 3002, an input / output device 3003, an output device 3004, an external storage device 3005, a portable recording medium driving device 3006 into which a portable recording medium 3009 is inserted, and a communication network 3007. These have a configuration in which they are connected to each other by a bus 3008. The configuration shown in the figure is an example of a computer that can implement the above system, and such a computer is not limited to this configuration.

  The CPU 3001 controls the entire computer. The memory 3002 is a memory such as a RAM that temporarily stores a program or data stored in the external storage device 3005 (or the portable recording medium 3009) when executing a program, updating data, or the like. The CUP 3001 performs overall control by reading the program into the memory 3002 and executing it.

  The input / output device 3003 detects an input operation performed by a user using a keyboard, a mouse, or the like, notifies the CPU 3001 of the detection result, and outputs data sent under the control of the CPU 3001 to a display device or a printing device.

The external storage device 3005 is, for example, a hard disk storage device. Mainly used for storing various data and programs.
The portable recording medium driving device 3006 accommodates a portable recording medium 3009 such as an optical disk, SDRAM, or Compact Flash (registered trademark), and has an auxiliary role for the external storage device 3005.

The communication interface 3007 is a device for connecting, for example, a LAN (local area network) or WAN (wide area network) communication line.
The system according to the first, second, or third embodiment is configured as shown in FIGS. 4, 5B, 13A, 13B, 15, 16, 17, 17, 19, and 21 for realizing the functions of the embodiments. This is realized by the CPU 3001 executing each control program corresponding to each operation flowchart of FIG. 22, FIG. 25, or FIG. The program may be recorded and distributed in, for example, the external storage device 3005 or the portable recording medium 3009, or may be acquired from the network by the communication interface 3007. The patent database 101 and rejection reason database 105 shown in FIGS. 1, 14, 24A, and 24B, the expression pattern dictionary 112, the word pair database 1302 shown in FIGS. 14, 24A, and 24B, and the like are, for example, an external storage device 3005. It is memorized and operated. In addition, the above-described work memory area is expanded on the memory 3002 as necessary.

Regarding the above first to third embodiments, the following additional notes are further disclosed.
(Appendix 1)
A device for assisting in understanding a word or a combination of words in a patent document,
A receiving unit that receives information for specifying a specification that is a target indicating a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. A document search unit indicating a reason for refusal to search text data of a document to be displayed based on information specifying the specification;
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words An extraction unit for extracting
An output unit for outputting a pair of character strings of the extracted paired words or word combinations;
A device for exemplifying a word or a combination of words in a patent document.
(Appendix 2)
The extraction unit includes:
String pattern matching processing and
Processing for position detection in strings,
Unnecessary character string deletion processing based on pattern matching with dependency character dictionary and dependency analysis,
The apparatus for exemplifying a word or a combination of words in the patent document according to Supplementary Note 1, wherein:
(Appendix 3)
The extraction unit includes:
When the character string of the pair expression pattern exists in the document notifying the reason for rejection, the application specific pattern and the cited document specific pattern are detected, and the paired character string is specified before or after these patterns. Or
If there is a character string in the order of the first character string, the character string pattern “corresponds to”, and the second character string following the character string pattern “in”, the first character string and the character string pattern Extracting a second character string as the paired character string pattern;
The example apparatus of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned.
(Appendix 4)
The process of deleting the unnecessary character string is as follows:
A morpheme analysis process that selects a morpheme string having the highest probability, using a dictionary of morphemes whose elements are parts of speech and character strings, and a table defining connection probabilities of parts of speech,
Clause separation processing for generating clauses by connecting morphemes according to clause connection rules;
It consists of dependency analysis processing that selects the dependency destination with the highest probability according to the connection probability of the clause and the constraint rule.
The example apparatus of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned.
(Appendix 5)
In the text data of the specification of the cited reference cited in the description of the document indicating the reason for refusal and the text data of the specification to be rejected, the pair of words or combinations of words By identifying the appearance position of the character string data corresponding to each constituting word or combination of words in the text data of each specification, each word constituting the paired word or combination of words or A specific example determination unit that determines which of the word combinations is a word indicating a specific example or a combination of words;
The example apparatus of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned.
(Appendix 6)
The specific example determination unit determines which one of the words or word combinations constituting the pair of words or word combinations that is the pair only by specifying the appearance position is a word or combination of words indicating a specific example. If the determination is not possible, the patent database storing the specification data with the character string corresponding to each word or combination of words constituting the pair of words or combinations of words is searched, and the claim of the specification Which word or word combination constitutes a specific example based on the number of occurrences of each word or word combination in the embodiment and in the embodiment, which word or word combination constitutes the paired word or word combination set To determine if it is a combination of
The example apparatus of the word in the patent document of Claim 5, or the combination of a word characterized by the above-mentioned.
(Appendix 7)
A device for assisting in understanding a word or a combination of words in a patent document,
An input unit for sequentially inputting text data of a specification to be used as a specific example of a word or a combination of words included in the patent document from a database storing the specification;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. A document search unit indicating a reason for refusal to search text data of a document to be displayed based on information specifying the specification;
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A character string pair of a word or a combination of words shown in contrast to the word or the combination of words described in the text data of the description of the specification to be a specific example of the combination An extractor for extracting;
A word or word combination pair database that stores a set of character strings of the extracted pairs of words or word combinations;
An example device for a word or a combination of words in a patent document characterized by comprising:
(Appendix 8)
An apparatus for assisting in understanding a word or a combination of words in a patent document using the word or word combination pair database according to appendix 7,
An instruction unit that receives information for specifying a specification that is a target indicating a specific example of a word or a combination of words included in the patent document;
By searching for the word or word combination contained in the word or word combination pair database in the patent document, the word or word combination included in the patent document is paired with the word or word combination character string. A search unit for searching for a word or a combination of words indicating a specific example of the combination;
An output unit for outputting a set of character strings of the searched word or word combination;
An example device for a word or a combination of words in a patent document characterized by comprising:
(Appendix 9)
A device to help understand words or combinations of words in patent documents,
Receiving information identifying a specification that is a target of a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. Search the text data of the document to be shown based on the information specifying the specification,
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words Extract
Outputting a set of character strings of the extracted paired words or word combinations;
A program for executing functions.
(Appendix 10)
The extraction process is as follows:
String pattern matching processing and
Processing for position detection in strings,
Unnecessary character string deletion processing based on pattern matching with dependency character dictionary and dependency analysis,
The program according to appendix 9, characterized by including:
(Appendix 11)
The extraction process is as follows:
When the character string of the pair expression pattern exists in the document notifying the reason for rejection, the application specific pattern and the cited document specific pattern are detected, and the paired character string is specified before or after these patterns. Or
If there is a character string in the order of the first character string, the character string pattern “corresponds to”, and the second character string following the character string pattern “in”, the first character string and the character string pattern Extracting a second character string as the paired character string pattern;
The program according to appendix 9, which is a process.
(Appendix 12)
The process of deleting the unnecessary character string is as follows:
A morpheme analysis process that selects a morpheme string having the highest probability, using a dictionary of morphemes whose elements are parts of speech and character strings, and a table defining connection probabilities of parts of speech,
Clause separation processing for generating clauses by connecting morphemes according to clause connection rules;
It consists of dependency analysis processing that selects the dependency destination with the highest probability according to the connection probability of the clause and the constraint rule.
The program according to appendix 9, characterized by:
(Appendix 13)
In the text data of the specification of the cited reference cited in the description of the document indicating the reason for refusal and the text data of the specification to be rejected, the pair of words or combinations of words By identifying the appearance position of the character string data corresponding to each constituting word or combination of words in the text data of each specification, each word constituting the paired word or combination of words or Further executing the function of determining which of the word combinations is a specific example word or a word combination;
The program according to appendix 9, characterized by:
(Appendix 14)
In the determination of the specific example, only by specifying the appearance position, which word or combination of words constituting the pair of words or combination of words is a word or combination of words indicating a specific example. If the determination is not possible, the patent database storing the specification data with the character string corresponding to each word or combination of words constituting the pair of words or combinations of words is searched, and the claim of the specification Which word or word combination constitutes a specific example based on the number of occurrences of each word or word combination in the embodiment and in the embodiment, which word or word combination constitutes the paired word or word combination set Is a process for determining whether the combination is
The program according to appendix 13, characterized by:
(Appendix 15)
A method for a computer to assist in understanding a word or combination of words in a patent document,
Receiving information identifying a specification that is a target of a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. Search the text data of the document to be shown based on the information specifying the specification,
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words Extract
Outputting a set of character strings of the extracted paired words or word combinations;
An example method of a word or a combination of words in a patent document, wherein the computer executes the above.
(Appendix 16)
The extraction process is as follows:
String pattern matching processing and
Processing for position detection in strings,
Unnecessary character string deletion processing based on pattern matching with dependency character dictionary and dependency analysis,
The method for illustrating a word or a combination of words in the patent document according to supplementary note 15, characterized by including:
(Appendix 17)
The extraction process is as follows:
When the character string of the pair expression pattern exists in the document notifying the reason for rejection, the application specific pattern and the cited document specific pattern are detected, and the paired character string is specified before or after these patterns. Or
If there is a character string in the order of the first character string, the character string pattern “corresponds to”, and the second character string following the character string pattern “in”, the first character string and the character string pattern Extracting a second character string as the paired character string pattern;
The method for illustrating a word or a combination of words in the patent document according to Supplementary Note 15, wherein the method is processing.
(Appendix 18)
The process of deleting the unnecessary character string is as follows:
A morpheme analysis process that selects a morpheme string having the highest probability, using a dictionary of morphemes whose elements are parts of speech and character strings, and a table defining connection probabilities of parts of speech,
Clause separation processing for generating clauses by connecting morphemes according to clause connection rules;
It consists of dependency analysis processing that selects the dependency destination with the highest probability according to the connection probability of the clause and the constraint rule.
The example of the word or the combination of words in the patent document of Additional remark 15 characterized by the above-mentioned.
(Appendix 19)
In the text data of the specification of the cited reference cited in the description of the document indicating the reason for refusal and the text data of the specification to be rejected, the pair of words or combinations of words By identifying the appearance position of the character string data corresponding to each constituting word or combination of words in the text data of each specification, each word constituting the paired word or combination of words or The computer further determines which of the word combinations is a specific example word or combination of words;
The example of the word or the combination of words in the patent document of Additional remark 15 characterized by the above-mentioned.
(Appendix 20)
In the determination of the specific example, only by specifying the appearance position, which word or combination of words constituting the pair of words or combination of words is a word or combination of words indicating a specific example. If the determination is not possible, the patent database storing the specification data with the character string corresponding to each word or combination of words constituting the pair of words or combinations of words is searched, and the claim of the specification Which word or word combination constitutes a specific example based on the number of occurrences of each word or word combination in the embodiment and in the embodiment, which word or word combination constitutes the paired word or word combination set Is a process for determining whether the combination is
An example method of a word or a combination of words in the patent document according to supplementary note 19, characterized by:
(Appendix 21)
A device for assisting in understanding a word or a combination of words in a patent document,
A receiving unit for receiving information specifying a specification including the patent document;
Indicates the reason for refusal by searching the text of the document indicating the reason for refusal based on the information specifying the specification from the reason for refusal database recording the document indicating the reason for refusal. A document search section;
From the text of the document indicating the reason for the retrieved rejection, the word or word combination character string described in the patent document and the document other than the specification are described and shown in contrast to the word or word combination. An extraction unit for extracting a character string of a word or a combination of words,
An output unit for outputting the extracted paired word or word combination data;
A device for exemplifying a word or a combination of words in a patent document.

DESCRIPTION OF SYMBOLS 101 Patent database 102 Patent document search part 103 Application number extraction part 104 Rejection reason notification search part 105 Rejection reason database 106 Word pair extraction part 107 Expression pattern dictionary 108 Specific example addition object specific part 109 Display part 201 Gazette data 202 Rejection reason 1401 Specific Example determination unit 1401-1 Patent document analysis unit 1401-2 Word appearance position specifying unit 1401-3 Patent document search hit number acquisition unit 1401-4 Determination unit 1402 Word pair database 2401 Word pair search unit

Claims (7)

An apparatus for assisting in understanding a word or a combination of words in a patent document,
A receiving unit that receives information for specifying a specification that is a target indicating a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. A document search unit indicating a reason for refusal to search text data of a document to be displayed based on information specifying the specification;
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words An extraction unit for extracting
An output unit for outputting a pair of character strings of the extracted paired words or word combinations;
A device for exemplifying a word or a combination of words in a patent document. The extraction unit includes:
When the character string of the pair expression pattern exists in the document notifying the reason for rejection, the application specific pattern and the cited document specific pattern are detected, and the paired character string is specified before or after these patterns. Or
If there is a character string in the order of the first character string, the character string pattern “corresponds to”, and the second character string following the character string pattern “in”, the first character string and the character string pattern Extracting a second character string as the paired character string pattern;
The example device of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned. The process of deleting the unnecessary character string is as follows:
A morpheme analysis process that selects a morpheme string having the highest probability, using a dictionary of morphemes whose elements are parts of speech and character strings, and a table defining connection probabilities of parts of speech,
Clause separation processing for generating clauses by connecting morphemes according to clause connection rules;
It consists of dependency analysis processing that selects the dependency destination with the highest probability according to the connection probability of the clause and the constraint rule.
The example device of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned. In the text data of the specification of the cited reference cited in the description of the document indicating the reason for refusal and the text data of the specification to be rejected, the pair of words or combinations of words By identifying the appearance position of the character string data corresponding to each constituting word or combination of words in the text data of each specification, each word constituting the paired word or combination of words or A specific example determination unit that determines which of the word combinations is a word indicating a specific example or a combination of words;
The example device of the word in the patent document of Claim 1, or the combination of a word characterized by the above-mentioned. A device for assisting in understanding a word or a combination of words in a patent document,
An input unit for sequentially inputting text data of a specification to be used as a specific example of a word or a combination of words included in the patent document from a database storing the specification;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. A document search unit indicating a reason for refusal to search text data of a document to be displayed based on information specifying the specification;
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A character string pair of a word or a combination of words shown in contrast to the word or the combination of words described in the text data of the description of the specification to be a specific example of the combination An extractor for extracting;
A word-pair database that stores a set of character strings of the extracted pairs of words or word combinations;
An example device for a word or a combination of words in a patent document characterized by comprising: A device to help understand words or combinations of words in patent documents,
Receiving information identifying a specification that is a target of a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. Search the text data of the document to be shown based on the information specifying the specification,
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words Extract
Outputting a set of character strings of the extracted paired words or word combinations;
A program for executing functions. A method for a computer to assist in understanding a word or combination of words in a patent document,
Receiving information identifying a specification that is a target of a specific example of a word or a combination of words included in the patent document;
The reason for refusal indicated for the specification subject to the specific example of the word or combination of words from the refusal reason database recording the document showing the reason for refusal indicated for the specification. Search the text data of the document to be shown based on the information specifying the specification,
A character string having the same counter expression pattern as the pre-stored counter expression pattern is detected from the text data of the document indicating the reason for the retrieved rejection, and is described in the patent document based on the character string of the counter expression pattern. By detecting a pair of character strings each including a character string that is displayed and a character string shown in contrast to the character string, and deleting an unnecessary character string from the pair of character strings, A pair of a character string of a word or a combination of words described in text data of a specification to be used as a specific example of a combination and a character string of a word or a combination of words shown in contrast to the word or the combination of words Extract
Outputting a set of character strings of the extracted paired words or word combinations;
An example method of a word or a combination of words in a patent document, wherein the computer executes the above.

Images